The Forest Service and BLM managers of the Northern Coast Range AMA, at the recommendation of the Provincial Advisory Committee, have committed to implementing a large-scale, long-term management comparison in the AMA. The purpose is to compare alternative plausible approaches to answering the following big-picture question:
 

How can we restore and maintain late-successional forest habitat and ensure the integrity of North Coast ecosystems and also provide socioeconomic outputs to local communities?

Many issues addressed by the Northwest Forest Plan are landscape in scale, and therefore can only be addressed with stand-scale studies via extrapolation and professional opinions (e.g. FEMAT). For example:

• Providing appropriate habitat for species that travel or live over large areas (like northern spotted owls, salmon, and large mammals)

• Maintaining potential for plants, animals, and materials to move between suitable habitat patches (like providing for movement of logs and sediment through stream systems)

• Maintaining an appropriate mix of forest ages and community types on the landscape (like providing grass balds and alder stands in addition to old-growth conifer forests)

• Managing cutting cycles and road networks over large watersheds

Everyone has a pretty clear idea what an old-growth conifer forest looks like, but there are few clear ideas about what a functional late-successional landscape looks like (e.g. how much alder? how much stream disturbance? how much early-successional habitat?).
 

What is a landscape-level learning design?

A landscape-level learning design incorporates elements of experimental design, including: addressing questions with well-defined treatments (types of management), initial similarity of experimental units (where the treatments will be done), replication of consistently-applied treatments, and random assignment of treatments to units. Similarity of units is an elusive goal in many ecological studies; for example, two stands may seem similar to someon studying trees, but may seem quite different to someone studying mushrooms or lichens. Similarity of features such as climate, ownership, geology, seral stage distribution, and road density may be adequate for landscape-scale comparisons, particularly if the design is well-replicated.

The "treatments" in this case are being called "strategies", because they include the entire suite of management actions that occurs on a landscape. For example, varying the amount and type of silviculture in a landscape usually means varying the amount and types of roads, recreational facilities, and available revenues. In terms of ecosystem restoration, most opinions (professional and otherwise) fall along a continuum of mostly passive (i.e. fix the worst problems and let nature fix the rest) to very active (i.e. we can actively move the system in the direction we want it). Most participants in the process to date have voiced support for comparing 3 or more strategies along this continuum.

The scale of the landscape comparison is a critical decision, which determines what questions can be addressed adequately. The sixth-field watershed scale (5-10,000 ac) appears to be acceptable to most AMA managers and partners. Given the importance of salmon in the area, there is support for including a watershed and fish-bearing stream in each unit, but units could be larger to encompass upland ridge systems as well.

Selecting replicated units is an important step, since all studies assume that units are initially similar before treatments are begun. Multi-variate analyses of attributes like % federal land, % mature forest, climate, etc. have been conducted on the AMA using 6^th field watersheds. The first step suggested similar areas, which management teams are now examining to select units for further analysis.
 

What are some landscape-level questions?

Landscape-level questions could examine the specific mechanisms affecting an element or topic of interest ("research" questions) or could examine the response of that element to overall conditions in a watershed ("monitoring" questions). Except for some response variables that can be measured relatively easily (for example, with satellite imagery), systematic or random sampling with field surveys will be required at some regular intervals (e.g. 10 years) or following important events (e.g. large floods or windstorms). The types of response variables could include:

• amount and distribution of late-successional forest (rate of development)
• amount, size, and distribution of snags and/or logs in forests and streams
• composition and structure of riparian forests
• size and frequency of different disturbances: fire, blowdown, insect/disease outbreak, landslide
• number of people and the types of activities they are involved in
• amount and value of commodities extracted, remaining, and changes over time
• amount and nature of conflicts over management and resource use
• amount and viability of species X (coho, murrelet, tailed frog, etc.)
 

Landscape-level research questions could specifically address the spatial pattern of features on the landscape. For example:

• How do mature stands within large blocks differ from mature stands in checkerboard ownership blocks?
• How are edge effects (like blowdown and microclimate) affected by different amounts and distributions of seral stages?
• How are source-sink watershed processes (like movement of landslides, debris flows, and wood input into streams) affected by position on the landscape and different road densities and harvest patterns?
 

Selected Ecosystem Questions

The questions below are those currently selected for guiding the landscape study design. They tend to fall in the "monitoring questions" category because these seem attainable given current resources and interest. These do not preclude application of landscape-scale research questions to the selected units in the future.

1. What are the short-term, mid-term, and long-term effects of the different management strategies on habitat suitability for selected late-successional forest-dependent species?

2. What are the effects of the different management strategies in riparian and upland forests on stream habitat structure, landslide events, and potential for large wood recruitment in local streams?

3. What are the effects of the different management strategies on the diversity of forest types across the landscape and implications for utility of habitat for selected species dependent on late-successional forest?

4. How do the different management strategies affect development of late-successional habitat and cooperation amongst landowners where federal lands are intermingled with private forest lands managed on short rotations?

5. What are the effects of the different management strategies on levels and types of commodity outputs, administrative feasibility of management, recreational uses, and meeting the social values of local communities?
 

What's next?

We are currently examining the results of the preliminary similarity analysis and using them to select specific landscape units for further, more detailed analysis. At the same time, we are refining the 3 landscape management strategies and narrowing down the specific management practices for recreation, riparian restoration, roads, special forest products, and silviculture that will be pursued under each strategy. There are still plenty of opportunities to include your perspectives-we hope to hear from you!
 
 

modified October 22, 1998